Organizational Learning: A Review of Sorne

Literatures
Mark Dodgson*

Abstract

Mark Dodgson Organizational learning is currently the focus of considerable attention, and it is
Science Policy addressed by a broad range of literatures. Organization theory, industrial econ-
Research Unit, omics, economic history, and business, management and innovation studies all
University of approach the question of how organizations learn. A number of branches of
Sussex, Brighton, psychology are also revealing on the issue. This paper assesses these various
U.K.
literatures by exarnining the insights they allow in three main areas: first, the goals
of organizational learning; second, the learning processes in organizations; and
third, the ways in which organizationallearning may be facilitated and impeded.
It contends that while the various literatures are revealing in particular aspects of
organizationallearning, a more complete understanding of its complexity requires
a multi-disciplinary approach. The contributions of the different approaches are
analyzed, and sorne are as are suggested where the transfer of analytical concepts
may improve understanding.

Introduction

Conceptions of organizational learning are ubiquitous. Since Weber,

organizational theory has addressed the issue, and as the recent (1991)
special edition of Organization Science on 'Organizational Learning'
shows, it remains a current concern. The lengthy history of the study of
organizational learning is matched by the range of academic disciplines
studying it. For example, economic historians have examined the import-
ance of learning in the development of new industries and technologies
(Rosenberg 1976), and of the development of formal Research and
Development (R&D) as institutionalized learning mechanisms (Mowery
1981). Learning is argued by industrial economists to affect productivity
(Arrow 1962), and industrial structures (Dosi 1988). Learning within
firms has been a feature of the theory of the firm since Cyert and March
(1963), and learning plays a central role in Teece et al. 's (1990) 'dynamic
capabilities' theory of strategic management. The relationship between
learning and innovation is increasingly examined at a strategic manage-
Organization ment level (Dodgson 1991a; Loveridge and Pitt 1990) and at a tactical
Studies 1993, 14/3: management level concerned with new product introduction (Imai et al.
375-394 1985; Maidique and Zirger 1985). The idea of 'learning curves' is widely
© 1993 EGOS used in management education, and by commercial management con-
0170--8406/93
0014-0016 $2.00 sultants such as the Boston Consultancy Group.
376 Mark Dodgson

A number of reasons can be suggested why the study of organizational

learning is currentIy so fashionable. First, the concept of the 'Iearning
organization' is gaining currency amongst large organizations as they
attempt to develop structures and systems which are more adaptable and
responsive to change. This development has been described and
inftuenced by the work of a number of management analysts, such as
Peters and Waterman (1982), Kanter (1989) and Senge (1990). It is
increasingly appreciated that learning is a key to competitiveness (Garratt
1987). Second, and partIy related, is the profound inftuence that rapid
technological change is having on organizations. The turbulence
engendered by technological change in products, processes and organiza-
tion increases the uncertainties facing firms and the confticts within them.
The complexity of the new product development process (Rothwell
1992), and shortening product life-cycles, the transformation of produc-
tion processes towards, for example, 'lean production' (Womack et al.
1990), and the growing use of computer-assisted organizational innova-
tions such as Just-in-Time delivery systems and Materials Requirement
Planning increases the need for firms to learn to do things in new, and
often radically different, ways.
A third reason can also be suggested. The concept of 'learning' has a
broad analytical value. This is shown in the breadth of academic disci-
plines using it. Contemporarily, however, this has increased in value, as
on the one hand, normative approaches, such as those found in much of
the management literature on the subject, seek to find a new language to
de al with the changed circumstances facing firms. On the other hand,
sorne academic analysis of organizations, for example, in economics,
attempts to progress beyond existing static views of organizations as
'bundles of resources'. Learning is a dynamic concept, and its use in
theory emphasizes the continually changing nature of organizations. Fur-
thermore, it is an integrative concept that can unify various Ievels of
analysis: individual, group, corporate, which is particularly helpful in
reviewing the cooperative and community nature of organizations. An
interdisciplinary approach to learning avoids the rather introspective and
parochial views seen in sorne of the literature on the subject (seen, for
example, in the recent edition of Organization Science).

What is Organizational Learning?

There is rarely agreement within disciplines as to what learning is, and

how it occurs (Fiol and Lyles 1985), let alone agreement between disci-
plines. Economists tend to view learning either as simple quantifiable
improvements in activities, or as sorne form of abstract and vaguely-
defined positive outcome. The management and business literature often
equates learning with sustainable comparative competitive efficiency, and
the innovation literature usually sees learning as promoting comparative
innovative efficiency. These various literatures tend to examine the out-
Organizational Learning: A Review of Sorne Literatures 377

comes of learning, rather than delve into what learning actually is and
how these outcomes are achieved. In contrast, it is a major concern of
organization theory and psychology to examine the processes of learning.
Learning, in the sense used here, relates to firms, and encompasses both
processes and outcomes. It can be described as the ways firms build,
supplement and organize knowledge and routines around their activities
and within their cultures, and adapt and develop organizational efficiency
by improving the use of the broad skills of their workforces. This broad
definition incorporates a number of assumptions:
- learning generally has positive consequences even though the outcomes
of learning may be negative, i.e. firms learn by making mistakes.
- although learning is based on individuals in the workforce, firms can
learn in toto. While emphasizing the role of human agency in learning,
corporate and group culture is influenced by individuallearning and can
assist the direction and use of that learning.
- learning occurs throughout all the activities of the firm, and, as will be
argued later, it occurs at different speeds and levels. Encouraging and
coordinating the variety of interactions in learning is a key organiza-
tional task.
Firms that purposefully construct structures and strategies so as to
enhance and maximize organizational learning have been designated
'learning organizations'. The characteristics of the learning company are
described by Pedler et al. (1989) who define it as 'an organization which
facilitates the learning of all its members and continually transforms
itself', and argue that it:
- has a climate in which individual members are encouraged to learn and
to develop their full poten ti al.
- extends this learning culture to include customers, suppliers and other
significant stakeholders.
- makes human resource development strategy central to business policy.
- continually undergoes a process of organizational transform-
ation.
While Pedler et al. argue that there is no blueprint for learning firms,
students of large Japanese firms see many similar characteristics (Dore
1973; Dore and Sako 1989; Sako 1992), and there is considerable
semblance with smaller, entrepreneurial high technology firms (Dodgson
1991b). The heavy emphasis in such companies on training and human
resource development to facilitate learning is matched by efforts to con-
sider the direction and effective utilization of learning activities.
Organizationallearning is used by Pedler et al. in a metaphorical sense:
the transformation of organizations is seen as being similar to individual
learning. The use of metaphor is effectively used by Morgan (1986) who
uses a number of them in his examination of organization. He does so as
' ... the use of metaphor implies a way of thinking and a way of seeing
that pervades how we understand our world gene rally' . This paper will
argue that individuals are the primary learning entity in firms, and it is
individuals which create organizational forms that enables learning in
378 Mark Dodgson

ways which facilitate organizational transformation. It will use organiza-

tional learning as a metaphor for individuallearning, and the contextua)
and internal stimuli to learning in people will be examined for similarities
in learning in organizations. The aim of doing this is to highlight the value
of research from the established organization theory and psychology
traditions, particularly concerning the process of learning for the emerg-
ing literature on learning in the innovationlmanagementleconomic fields.
Concomitantly, this emerging literature is revealing, particularly concern-
ing the aims of technologicallearning, and in sorne of the factors facilitat-
ing and impeding learning for more established bodies of work.

The Goals of Organizational Learning

Before analyzing the processes of organizational learning, the question

should be posed: why do firms learn? Essentially, learning can be seen to
have occurred when organizations perform in changed and better ways.
The goals of learning are useful outcomes. Common explanations of the
need to learn is the requirement for adaptation and improved efficiency in
times of change. Psychologists, for example, see learning as the highest
form of adaptation, raising the probability of survival in changing
environments, and various other approaches also emphasize the need for
adaptation. While organizational theory often assumes learning to be
stimulated by the need for organizational adjustment in response to sorne
rather ilI-defined external stimulus, the management and innovation
literature is much more clear. Learning is seen as a purposive quest to
retain and improve competitiveness, productivity, and innovativeness in
uncertain technological and market circumstances. The greater the
uncertainties, the greater the need for learning. Amongst other factors
stimulating environmental uncertainty and learning, two are particularly
important at present: responses to technological change, and Western
companies' response to the competitiveness of alternative forms of
industrial organization (particularly J apanese).
Freeman and Perez's (1988) theory of changing 'techno-economic
paradigms' describes the way that profound changes in technologies,
most recently seen in information technology, cause considerable
environmental turbulence for firms and other institutions as they attempt
to respond to, as well as being part of, these radical changes. The goals of
learning in these circumstances can therefore be seen as a response to the
need for adjustment in times of great uncertainty. This theoretical
analysis has sorne support in the more empirically-based analysis of Pavitt
(1991), who argues that the strategies of large innovative firms are
determined in part by attempts to learn in highly uncertain conditions. In
an uncertain and rapidly developing new technology, biotechnology,
Dodgson (1991b) argues that it is the differential ability to learn quickly
about technological opportunities that has been responsible for the
changing pattern of competitive relationships between large and small
Organizational Learning: A Review of Sorne Literatures 379

firms which have been so important to the development of the

technology .
Leaming is a key feature in the process by which firms accumulate tech-
nology in order to compete. This is seen particularly clearly in Japanese
firms, whose industrial organization is increasingly seen as a model to be
replicated (Marceau 1992). Japanese firms place particular emphasis on
leaming (Pucik 1988b). According to Imai et al. (1985), Japanese firms
possess 'an almost fanatical devotion towards leaming - both within
organizational membership and with outside members of the inter-
organizational network. To them, leaming is something that takes place
continuously in a highly adaptive and interactive manner' (1985: 353). It
is this ability to leam which is argued to explain the success of Japanese
firms in continually introducing innovative products. The competitive
success of Japanese firms based upon this high level of innovation has
increased the need for Westem firms to adapt their organization, and in
industries such as automobiles, to attempt to emulate them (Womack et
al. 1990). Firms are leaming how competitor firms leam.
The efficiency goals of leaming are commonly equated with productivity,
the major focus of interest of industrial economists. For example, pro-
ductivity is argued to be assisted through 'Ieaming by doing' (Arrow
1962). This has been common parlance amongst economists in explaining
how manufacturing-unit costs reduce over time, or with experience.
Sorne of the more simplistic assumptions of this approach are highlighted
by Bell and Scott-Kemmis (1990) who argue that leaming by doing is
often assumed to be passive, automatic and costless, yet it is none of these
things. Adler (1990) provides a more sophisticated approach to manufac-
turing productivity improvements and argues that this is a feature of the
accumulation of knowledge and leaming across the development/
manufacturing interface, sharing between the primary location and plants
started up later, and continued sharing of knowledge amongst plants. In
addition Rosenberg (1976) describes 'leaming by using' based on the
wealth of experience derived from using new products and
processes.
In general, however, most economists' contribution to our understanding
of organizational leaming is limited to descriptive analyses of the out-
comes of cumulative experience. There are sorne exceptions. Develop-
ment economics, for example, has considered the processes by which
firms in developing countries leam (BellI984), and recently, economists
such as Cohen and Levinthal (1989) have begun to examine leaming in
R&D more broadly. They argue that leaming occurs not only in the focus
of R&D, but also in the process itself ... 'while R&D obviously gener-
ates innovations, it also develops the firm's ability to identify, assimilate,
and exploit knowledge from the environment ... a firm's "leaming" or
"absorptive" capacity' (1989: 569). This article provides one of the first
examples on the subject in the mainstream economics literature to con-
sider the outcome of R&D as enhanced organizational capability, as well
as new information.
380 Mark Dodgson

The complexity in the goals of learning can be seen in its occasionally

conftictual nature. Underpinning psychological theories on learning is the
assumption that conftict (caused, for example, by error or contrary
evidence) is an essential condition for learning, which acts as a motor
driving the learning process. This has sorne resonance in sorne of the
management and innovation literature which contends that although
learning may be motivated by attempts to improve innovativeness/adap-
tability or productivity/efficiency, these may be conftictual. Clark et al.
(1987) refer to the 'productivity dilemma' of a conftict between innova-
tion and productivity, change and experience, (what economists
sometimes refer to as the tensions between 'dynamic' and 'allocative'
efficiencies; and organization theorists, the tensions between 'explora-
tion' and 'exploitation'). They argue that an 'approach to resolving the
contradiction between efficiency and innovation is contained in the COIl-
cept of "learning", which highlights the dynamic interaction of the two
dimensions'. This begins to describe the dialectic in organizationallearn-
ing, mirroring that found in psychology. Piaget's dialectic in develop-
mental psychology, for example, assumes that existing cognitive
structures and the knowledge they engender (thesis) are continually chal-
lenged by new knowledge which does not fit (antithesis), and are eventu-
ally reorganized so that new knowledge is better integrated (synthesis and
new thesis). In the analogy, productivity and 'exploitation' are thus chal-
lenged by confticting innovation, which eventually leads to better
productivity.
By building on the metaphor of individuallearning, conftict, and hence
learning, can be seen as inevitable in organizations as it is in individuals.
It is a natural state. Organizationallearning is as natural as learning in
individuals as they attempt to adjust and survive in an uncertain and
competitive world. Furthermore, just as psychologists distinguish various
'levels' of learning, progressing from biological or adaptive learning, the
'Iearning organization' can be distinguished as one that moves beyond
this 'natural' learning, and whose goals are to thrive by systematically
using its learning to progress beyond mere adaptation. It is an organiza-
tion which attempts to develop what psychologists see in individuals as
higher level, constructive or generative mental functions, and is reftected
in strategies and structures purposefully being developed to facilitate and
coordinate learning in rapidly changing and conftictual circumstances.

The Processes of Learning

Sorne of the complexity within the processes of organizational learning

can be analyzed by continuing with the metaphor of individuallearning.
Corsini (1987) sees individual learning as involving five kinds of learned
capabilities. Verbal knowledge (declarative knowledge) ranges from
isolated 'facts' to bodies of organized information. Intellectual skills (pro-
cedural knowledge) enable the individual to demonstrate the application
Organizational Learning: A Review of Sorne Literatures 381

of concepts and rules to specific instances. Cognitive strategies involve a

number of processes such as perceiving, encoding, retrieving and think-
ing; they can be problem-solving, and can control and modify other
cognitive processes of leaming and memory such as attention, encoding
and retrieval. Attitudes are 'leamed states that inftuence the choices of
personal action the individual makes towards persons, objects or events'.
Motor skills are smoothly timed muscular movements enabling pro-
cedures to be undertaken precisely (this latter category will not be con-
sidered here, although there are obvious similarities between it and
'leaming by doing').
The process of building deelarative knowledge is akin to the economists'
view of 'search' activities. It ineludes all the methods by which firms glean
information: for example, through R&D, education and training, recruit-
ment, and techniques such as patent watching and bibliometrics. The
management and innovation literature is voluminous in this area.
However, it is in the process of building procedural knowledge, cognitive
strategies and attitudes that the major organizational challenge lies. As
Corsini suggests, it is not enough to 'know what', it is also essential to
'know how'.
A number of studies into leaming from a management studies perspective
similarly distinguish various types and levels of leaming. Fiol and Lyles
(1985), for example, distinguish higher and lower level leaming. Senge
(1990) differentiates generative from adaptive leaming. Dodgson (1991a)
separates strategic from tactical leaming. From an organization theory
perspective, Argyris and Schon (1978) develop a three-fold typology of
leaming which they describe as single-loop, double-Ioop and deutero-
leaming, which has many similarities to approaches to individual
leaming.
'Organizationallearning involves the detection and correction of error. When the
error detected and corrected permits the organization to carry on its present
policies or achieve its present objectives, then that error-detection-and-correction
process is single-loop learning. Double-loop learning occurs when error is detec-
ted and corrected in ways that involve the modification of an organization's
underlying norrns, policies and objectives.' (1978: 3)

Organizations need to leam how to carry out single and double-Ioop

leaming according to Argyris and Schon, who call this deutero-Ieaming.

'When an organization engages in deutero-learning its members learn about

previous contexts for learning. They reflect on and inquire into previous episodes
of organizational learning, or failure to learn. They discover what they did that
facilitated or inhibited learning, they invent new strategies for learning, they
produce these strategies, and they evaluate and generalize what they have produ-
ced.' (1978: 4)

The higher level 'double-Ioop' and 'deutero' leaming compares with

Corsini's (1987) categories of cognitive strategies and attitudes. How
these translate from individuals to organizations critically depends upoo
382 Mark Dodgson

organizational culture. For Schein (1985), culture is a learned product of

group experience. He defines it as ...

'basic assumptions and beliefs that are shared by members of an organization, that
operate unconsciously, and that define in a basic "taken for granted" fashion an
organization's view of itself and its environment. These assumptions and beliefs
are learned responses to a group's problems of survival in its external enVÍron-
ment and its problems of internal integration.' (1985: 6)

The relationship between individual and collective learning has received a

great deal of attention from within the organization theory perspective.
Throughout the various literatures, firms' learning is commonly argued to
be more than the sum of the parts of their employees' learning. Shared
norms and values are agreed to be indicative of organizational rather than
individuallearning. According to Hedberg (1981), for example:

'Although organizationallearning occurs through individuals, it would be a mis-

take to conclude that organizational learning is nothing but the cumulative result
of their members' learning. Organizations do not have brains, but they have
cognitive systems and memories .. . Members come and go, and leadership
changes, but organizations' memories preserve certain behaviours, mental maps,
norms, and values over time.' (1981: 3)

As Schein argues, it is the internal integration of individuals within a

shared culture that facilitates learning, and this provides a major
organizational challenge. Particularly important in the definition of this
culture is technology. Schein argues that 'Occupations typically build
their practices, values and basic self image around their underlying tech-
nology'. Technologically successful organizations, he argues, tend to base
their self-image around their technology. The role that organizational
culture plays in merging individual and collective learning warrants fur-
ther interdisciplinary research.
There are a number of approaches within the management/innovationJ
economics literatures which refer to the centrality of identifiable and
cohesive bodies of, often technologically related, knowledge and
behaviour. Just as psychologists such as Corsini suggest that learned
capabilities are dependent upon certain prior conditions, such as the
presence of existing knowledge, and internal and environmental stimuli,
so too do these other literatures. Three main concepts used to analyze the
activities and behaviour of firms which direct and inform learning in a
collective rather than individual sense are described here: 'knowledge-
base', 'firm-specific competences', and 'routines'. These are conceptually
different, although when discussing the impact and importance of learn-
ing for them, there are sorne similarities.
The term knowledge base is used to analyze the form of knowledge and
the focus of its accumulation (Metcalfe and Gibbons 1989). Organiza-
tional uniqueness is defined by knowledge bases and the processes of
acquisition, articulation and enhancement of the knowledge over which it
has control. There are sorne similarities between this approach and firm-
Organizational Learning: A Review 01 Sorne Literatures 383

specific competences (Pavitt 1991); 'firm-specific capabilities' (Teece et al.

1990) and 'core competencies' (Prahalad and Hamel 1990). These
approaches also argue the uniqueness of firms' knowledge and learning.
For Pavitt, technological development is firm specific, cumulative, and
differentiated. Teece et al.'s theory is based on the 'mechanisms by which
firms accumulate and dissipate new skills and capabilities, and the forces
which Iimit the rate and direction of this process·. Prahalad and Hamel
argue that competitiveness in the 1990s depends on firms nurturing their
core competencies, which they describe as the 'collective learning of the
organization', and involves: 'harmonizing streams of technology; the
organization of work; delivery of value; and, communication, involve-
ment, and a deep commitment to working across organizational
boundaries' .
For Nelson and Winter (1982), it,is the establishment of routines which
operationalize organizations' memories and knowledge bases. Routines
in organizations are argued to inelude:
' ... the forms, rules, procedures, conventions, strategies, and technologies
around which organizations are constructed and through which they operate. It
also ineludes the structure of beliefs, frameworks, paradigms, codes, cultures and
knowledge that buttress, elaborate, and contradict the formal routine.' (Levitt
and March 1988: 320)

Such a broad definition has limited value, and Nelson and Winter them-
selves use the term 'routine' flexibly. However, it has sorne value in that
the concept of routine implies organizational action. That is, returning to
Corsini's distinction between deelarative and procedural knowledge, it is
not only what a firm knows or what skills it possesses, but how it uses
these which is important.
Conflating the terminology used aboye, we can equate single-loop learn-
ing with those activities which add to the knowledge base or firm-specific
competences or routines of the firm without alteriog the nature of their
activities. It can be seen as analogous to the development of Corsini's
verbal knowledge. Double-loop learning involves changing the firms'
knowledge base, firm-specific competences and routines, and is anal-
ogous to intellectual skills. double-Ioop and deutero-Iearning involves
consideration of why and how to change, and hence the analogy with the
development of cognitive strategies and attitudes.
These three concepts usefully illustrate sorne of tbe forms of collective
learning and its importance for firms. The nature of the knowledge base
or 'firm-specific competence' is individual to particular firms and is a
crucial factor affecting their competitiveness. Collective learning is
dynamic, but the way that it develops is constrained by existing ways of
doing things, know-how and routines. Within the economicslinnovation.
management literature there are broad analyses of the dynamic of collec-
tive learning in technology, describing the way it has its own trajectories
(Dosi 1982) and how firms' learning is 'path dependent' (Dosi 1988). As
Pavitt (1991) argues:
384 Mark Dodgson

' ... the range of possible choices about both product and process technologies
open to the firm depends on its accumulated competence ... the improvement of
these competences requires continuous and collective leaming.' (1991: 42)

The centrality of technological and knowledge accumulation for the com-

petitiveness of firms is something that is appreciated comparatively more
keenly by recent management and innovation literatures. However, aH
these approaches take too little account of the importance of individual
human agency and the complexity and problems involved in leaming.
Routines, for example, are argued to be independent of the individuals
who operate within them and use them, and are capable of surviving
considerable tumover in individual actors. Yet routines are responses
chosen by individuals. Bandura (1977), for example, refers to the promi-
nent role of 'self regulatory capacities' of individuals, and the way self-
inftuence partly determines which actions one performs. 'By arranging
environmental inducements, generating cognitive supports and producing
consequences for their own actions, people are able to exercise sorne
measure of control over their own behaviour' (1977: 13).
Furthermore, these analyses of coHective leaming tend to assume
uniformity in leaming capabilities within firms. In complex organizations
many different leaming processes can proceed at the same time in dif-
ferent directions and at different speeds. Marengo (1992) shows that even
when aH the members of an organization are characterized by a given and
constant leaming process, the way in which knowledge, information and
communication ftows are distributed in the system can give rise to very
different pattems of organizationalleaming. This highlights the import-
ant role of coordinating mechanisms for leaming considered in the next
section.
Tuming now to individual leaming, a number of literatures point to the
importance of key individuals in organizational leaming. Included are
those which act to bring information into the firm from outside - what
the social psychology literature knows as 'boundary spanners' (Michael
1973), and the innovation literature denotes as 'technological
gatekeepers' (AHen 1977). However, this says Httle of the factors, apart
from their existence, which affect their own, and their organization's,
leaming.
Individuals' leaming is constrained by their ability to interpret complex
reality: Simon's (1957) bounded rationality problem. Furthermore, their
leaming is sociaHy constructed inasmuch as what is leamed is profoundly
connected to the conditions in which it is leamed. Simon (1991) argues
that although aHleaming takes place inside individual human heads ...

'What an individual leams in an organization is very much dependent on what is

already known to (or believed by) other members of the organization and what
kinds of information are present in the organizational environment.' (1991: 125)

Schein (1985) refers to the need of organizations to socialize new entrants

with its pivotal or central assumptions, and describes the importance of
Organizational Learning: A Review of Sorne Literatures 385

culture 'to be taught to new members as the correct way to perceive,

think and feel in relation to those problems' .
Learning is not only socially constructed, specific to particular firm and
group cultures, and conflictual, it can also be conservative and unreliable.
March et al. (1991) argue that what is learned from any particular kind of
experience can vary substantially across time and across organizations.
They describe the importance of experience, particularly of 'critical
incidents' in shaping learning. However, ...

'(b)ecause of the ambiguities associated with any single incident, responses and
interpretations tend to be adopted more as a result of their temporal proximity;
cognitive availability, or political convenience than by virtue of their obvious
validity.' (March et al. 1991: 7)

They further contend that the learning process is generally conservative

and sustains existing structures of belief (with obvious parallels with
Kuhn's 'normal science'). They distinguish 'reliable' learning processes
- which enable an organization to develop 'common understandings of
its experience and makes its interpretations public, stable and shared' -
and 'valid' learning processes which allow an organization to 'under-
stand, predict and control its environment'. Organizations need both
reliable and valid learning processes, although the two can conflicto

'Stable, shared knowledge interferes with the discovery of contrary experience

from which valid leaming arises, and the exploration of novel ideas interferes with
the reliable maintenance of sharing of interpretations.' (March et al. 1991: 6)

March (1991) continues this examination of tensions induced by learning,

and sorne of the resultant consequences, in his examination of 'explora-
tion' and 'exploitation' in organizationallearning ...

'The essence of exploitation is the refinement and extension of existing com-

petences, technologies and paradigms. Its retums are positive, proximate, and
predictable. The essence of exploration is experimentation with new altematives.
Its retums are uncertain, distant and often negative. Thus, the distance in time
and space between the locus of leaming and the locus for the realization of retums
is generally greater in the case of exploration than in the case of exploitation, as is
the uncertainty.
Such features of the context of adaptation lead to a tendency to substitute exploi-
tation of known altematiges for the exploration of unknown ones.' (March 1991:
85)

Another problem involved in learning is 'unlearning', or forgetting past

behaviour which is redundant or unsuccessful. This is very important for
firms (Clark et al. 1987), and is a crucial feature of Hedberg's (1981)
analysis:

'Knowledge grows, and simultaneously it becomes obsolete as reality changes.

Understanding involves both leaming new knowledge and discarding obsolete
and misleading knowledge. The discarding activity - unleaming - is as import-
386 Mark Dodgson
---------------------------------_. __._--
ant a part of understanding as is adding new knowledge. In fact, it seems as if slow
unleaming is a crucial weakness of many organizations.' (1981: 3)

A wide range of sources of firm learning have been suggested. R&D is a

key source, and, additionally, from a development economics perspec-
tive, Hobday (1990) describes: learning via joint ventures; by installing
capital goods, training, hiring key individuals, reverse engineering,
designing, imitating inward investors; and through assembly and mass
manufacturing, exporting and investment abroad. The implications of
these analyses is that leaming occurs in a number of functions within the
firm: research, development, design, engineering, manufacturing, and
marketing, and externally, and they suggest how important it can be to
elicit and utilize information from aH these sources. The sources of learn-
ing may vary over time with industry, technology and product life cycles,
according to the aims of the firm (Mody 1990). Thus with the initial stages
of industrial growth, and technology and product development and diffu-
sion, learning may focus on overcoming uncertainties. In later stages, the
focus of learning may be attempts to achieve benefits of scale. Beyond
this, efforts may focus on regeneration or de-maturity.
Particularly important for innovation in firms is learning from customers
and users (von Hippel 1988; Stinchcombe 1990; Rosenberg 1982). This
'external' learning provides an important learning process. Bandura
(1977) emphasizes the way people learn by watching before they perform,
the way they can profit from the successes and mistakes of others as well
as from their own experiences. 'The capacity to learn by observation
enables people to acquire large, integrated patterns of behaviour without
having to form them gradually by tedious trial and error.' (1977:
12)
The innovation literature describes this when it distinguishes between
'leader' and 'foHower' strategies (Freeman 1982), with sorne firms bear-
ing high risks and costs in the expectation of super profits and others
content to entertain less risky and profitable strategies. Recently, sorne of
the research in this tradition which examines collaboration between
firms, in its various forms - joint ventures; strategic allianees; R&D
contracts - sees it as an opportunity to learn from partners (Ciborra
1991; Dodgson 1993b). Whipp et al. (1990) describe an example of the
way managers and engineers in eompeting firms with shared values and
aims learn direetly from one another. There would, however, be con-
siderable value in wider researeh to examine the ways in which vicarious
learning manifests itself in organizational ehange. Such research would
have significance for better understanding of organizational behaviour in
a variety of relationships - eustomerlsupplier and competitive 0.- and
may have sorne import for mergers and acquisitions.
When considering the processes oí organizationallearning, the metaphor
oí individuallearning adds to our 'way of seeing' in Morgan's sense. The
emphasis in mueh of the managementlinnovationleeonomics literature is
that 'history matters', and what a firm can do in the future is strongly
Organizational Learning: A Review of Sorne Literatures 387

inftuenced by its past and its collective leaming. By examining the

inftuences individuals have in creating their own cultures (and histories in
this sense), a less deterministic view can be taken. Furthermore, as ideas
such as vicarious leaming and different levels of leaming are introduced,
sorne of the analytical tools for conceptualizing organizational transform-
ation are presented.

Factors Facilitating and Impeding Learning

Throughout the aboye discussion the importance of environmental

change as a stimulus to leaming has continually been emphasized. It has
been argued that the greater the uncertainty in the environment, the
greater the need for leaming. However, on the question of adaptation,
for example, sorne approaches which see it primarily as a response to
environmental change suffer from the reductionist tendency of
behaviourism. Organizationalleaming cannot be created and eradicated
by varying extemal stimuli. Organizations, and the forms of collective
and individualleaming within them, importantly affect leaming processes
and outcomes. Indeed, the role of human agency and individual goals
such as the drive for self-actualization are almost completely ignored in
many accounts of organizational adaptability.
Organizationalleaming is stimulated both by environmental changes and
intemal factors in a complex and iterative manner. Retuming to the
individualleaming metaphor, Bandura's (1977) Social Leaming Theory,
for example, argues that ...

'. . . people are neither driven by inner forces nor buffeted by environmental
stimuli. Rather, psychological functioning is explained in terms of a continuous
reciprocal interaction of personal and environmental determinants.' (1977: 12)

By using the individual leaming metaphor, the complexity in the factors

that encourage and constrain leaming begin to be delineated. In addition
to the levels of leaming, and the complexity in the collective/individual
leaming relationship described in the previous section, the situation is
further complicated by the environmentlorganization interface. One fac-
tor which can assist organizations to deal with this particular problem is
strategy, and this is where the managementlinnovationleconomics
literature is particularly revealing.
Firms purposefully adopt structures and strategies to encourage leaming.
They are not totally reactive, and can proactively seek to inftuence the
environment in which they leam. The relationship between strategy and
structure, and the role of strategy acting as an intermediary filter between
the environment and the organization has long be en a focus interest in the
management literature (Ansoff 1968; Chandler 1966; Snow and Miles
1983). However, recent insights into the challenge in creating organiza-
tional forms conducive to leaming are provided by contemporary econ-
388 Mark Dodgson

omic theory of the firm (Marengo 1992). This builds upon Coase (1937)
and argues that the nature of the firm is to be found in the mechanisms
(economic, social, political and psychological) which the firm itself
implements in order to achieve the necessary coordination amongst the
actions of its individual members. The firm is seen as a coordinating
institution, and learning is one of the activities which needs coordination,
and the mechanisms used to achieve such coordination playa central role
in shaping the organizational learning process and determining its
outcome.
Viewing the role of firms as coordinated learning institutions leads to a
consideration of the role of organizational structure and strategy. Com-
plex organizations are characterized by a multiplicity of learning proces-
ses: each individual and each group within the organization have their
own knowledge base and their own learning capabilities. The structure of
the organization defines the way in which these processes interact
(Chandler 1990), and gives rise to the organizational learning process
resulting from these interactions. Aoki (1988) describes the way in which
the information structures and dense communication ftows encouraged by
Japanese firms encourage employees' integrative learning. Dore (1986)
describes the importance of culture in inftuencing these structures. In this
regard corporate strategy to encourage learning has its primary focus on
creating such organizational structures and cultures. A feature of such a
strategy and structure ineludes consideration of the way firms may benefit
from the diversity and heterogeneity of learning. As March (1991)
argues, there is value for organizations in having fast- and slow-Iearning
individuals together. Although this creates complex problems of
coordination, it also encourages the scope of organizational learning.
Furthermore, it raises the critical question of incentives to learning (Aoki
1988), particularly the problem of designing incentive structures in such a
way as to make confticting and heterogenous learning compatible.
Another question relates to the way interorganizational learning is
facilitated by high levels of trust between firms (Sako 1992; Dodgson
1993a).
A feature of learning strategies is, of course, the level and direction of
resources devoted to learning. The R&D departments of firms provide a
major so urce of learning in an activity which is central to their continuing
existence and prosperity. The development of formal R&D structures in
firms has been argued to be a response to the need to learn effectively
about developing science and technology (Mowery 1981). It is the R&D
departments of firms that provide the major vehiele for learning about
new technological developments. As a major source of learning is R&D ,
the size and focus of R&D budgets are likely to be a primary factor
encouraging and constraining learning. This is obviously affected by deci-
sions made by managers, and there is an extensive literature on the
establishment of R&D budgets and project selection (Twiss 1986;
Dumbleton 1986), and on organizational questions of obtaining creative
efficiency in R&D (Hull 1988; Kolodny 1980). Learning is a costly pro-
Organizational Learning: A Review of Sorne Literatures 389

cess (Cohen and Levinthal 1989). However, as Pucik (1988a) points out,
traditional planning systems cannot assign a financial value to learning
activities, and hence they usually go unfunded. What is more, the costs of
learning are immediate and the benefits long-termo There is little con-
sideration within the managementlinnovationleconomics literature of the
actual and opportunity costs of learning. Furthermore, given the
centralíty of R&D as an organizational learning mechanism, there is a
surprising paucity of research across all traditions on the broad question
of learning in R&D, its promotion and funding, and the subsequent
diffusion of learning throughout the organization. The way that the pro-
cesses and outcomes of learning have been facilitated by recent technolo-
gies, such as multi-media in communications, information and training
provision, provide fertile ground for future research. Such research
would necessarily need to be interdisciplinary and should in elude the
range of insights from organization theory, which generally have been
comparatively slow in addressing these issues.
The above approaches valuably describe sorne important sources of
learning, and have implications for practical issues such as resource allo-
cation. However, the difficulties in organizational learning remain, and
these are often underestimated in the managementlinnovation/economic
perspectives. Argyris and Schon's (1978) study found that most organiza-
tions do quite well in single-loop learning, but have great difficulties in
double-Ioop learning. They could find no example of organizations which
learned in a deutero fashion. Their general contention is that organiza-
tions ordinarily fail to learn on a higher level. One reason for this is
because of what they describe as inhibitory loops. Primary inhibitory
learning loops are a self-reinforcing cyele in which errors in action pro-
voke individuals to behaviours which reinforce those errors. Secondary
inhibitory loops are group and inter-group dynamics which enforce condi-
tions for error (ambiguity, vagueness, etc.). They contend that organiza-
tions tend to create learning systems that inhibit double-Ioop learning,
calling into question their norms, objectives, and basic policies.
Morgan (1986) also analyzes learning inhibitors. He describes how
departmental structures focus the attention of their members on
parochial rather than organization-wide problems; how systems of
accountability frequently foster defensiveness in attitudes; and how, as
Argyris and Schon argue, there is a gap between actors' rationalized
statements of what they do and what actually occurs.
Although the problems of learning in organizations, such as the obstaeles
to unlearning, are considerable, learning, of course, can and does occur.
Argyris and Schon do not rule out the possibility of higher level double-
loop and deutero learning. Japanese organizations are argued to be cap-
able of a continuous process of learning, to their distinct advantage (Imai
et al. 1985). It is here that sorne of the research within an economics/
managementlinnovation perspective has implications for organization
theory. A number of major companies have transformed themselves.
Examples would be IBM, changing from a traditional 'business machines'
390 Mark Dodgson

company, and lel using science and technology to move out of bulk
chemicals. Without denigrating or underestimating the extent of the
problems firms face when attempting to learn - a factor often over-
looked by the management and business/economics/innovations
literatures - firms manage, in practice, to move in radical new directions
through higher level learning. Researchers in the organization theory
tradition may find value in examining the rapid and extensive technologi-
cal transformations undergone within companies, and the role of distinct
strategies in shaping their learning environment, discussed in depth in
these other literatures.

Conclusions

This review of some of the literature on organizational learning has

argued that the strengths of analyses within an organization theory
perspective, when supplemented by some concepts from psychology, líe
in comprehension of the process and problems of learning. The strengths
of the economics/management and businesslinnovations approaches lie in
motives and sources of leaming. It argues that a primary reason why firms
leam is to deal with uncertainty in their markets and technologies, and
while R&D is an important source of learning, leaming occurs
throughout the activities of the firmo These approaches valuably highlight
the contemporary significance of the accumulation of learning for com-
petitiveness, and the proactive role of strategy in its stimulation. Much of
its analysis of learning is, however, limited to its outcomes, and it ignores
or underestimates the problems and complexities in the processes of
leaming. Use of the individualleaming metaphor shows how organiza-
tionalleaming is problematic - it may be conftictual and conservative -
and why unleaming and higher-Ievel leaming is difficult. Together, the
literatures reviewed contribute to the understanding of the complexity of
factors that encourage and restrict leaming.
The aim of this paper has been to illustrate the potential overlaps and
synergies between the various approaches to leaming; a concept which all
the literatures reviewed agree has considerable analytical power. It has
not attempted fully to integrate the approaches in any systematic manner.
This may be an interesting thing to do. However, while there are areas of
agreement, there remains a great disparity in the fundamental underlying
assumptions of the different approaches, such as the differing focuses on
outcomes and processes.
Syothesis of these differiog approaches, aod the development of an inter-
disciplinary perspective, will have particular value for the continuing
study of 'leaming organizations'. The characteristics of such organiza-
tions are only just beginning to be delineated. The competitive value of
leaming is very high, and research ioto the area is likely to increase in
topicalíty. This review points to the value of using some of the concepts
aod sub-concepts in the various literatures in any future analysis.
 Organizational Learning: A Review of Sorne Literatures 391

Examination of concepts such as vicarious learning, differing levels of

learning, and questions of the formulation and maintenance of culture
and its relationship with individual learning warrant broader analytical
attention. Additionally, it suggests that a deeper examination of techno-
logicallearning, and sorne of the coordinating mechanisms that firms use
through their strategies and structures, and, for example, the way they
use incentives to learning, would be very valuable. Such enquiry would
have both normative and theoretical value. For example, how do
managers ensure the compatibility of individual learning with organiza-
tional aims? What are the aims and characteristics of learning strategies
(what psychologists call 'metacognition ')? How, once it has been
generated or accessed, is learning diffused throughout organizations and
shared by organizational means, such as personnel transfer, or techno-
logical means, such as electronic databases. Such enquiry would have
considerable significance for the study of contemporary organizational
behaviour.

Note *The author wishes to thank Martin Bell, Jane Millar, Luigi Marengo, Keith Pavitt and two
anonymous referees for their very helpful eomments on this papero

References Adler, P. Bell, M., and D. Seott-Kemmis

1990 'Shared Learning'. Management
Science 36/8: 938-957.
AlIen, T.
1977 Managing the flow 01 technology.
Boston, Mass.: MIT Press.
Argyris, C., and D. Sehon
1978 Organizational learning. London:
Addison-Wesley .
Arrow, K.
1962 'The implieations of learning by
doing'. Review 01 Economic Studies
29: 166-170.
Bandura, A.
1977 Social learning theory. Englewood
Cliffs NJ.: Prentiee-Hall.
Bell, M.
1984 '''Learning'' and the aeeumulation
of industrial teehnologieal eapaeity
in developing eountries' in Techno-
logical capacity in the Third World.
K. King and M. Fransman (eds.),
189-209. London: Maemillan.
1990 'The mythology of learning-by-
doing in World War 11 airrrame and
ship produetion' . Seienee Poliey
Researeh Unit, University of
Sussex.
Chandler, A.
1966 Strategy and structure. Cambridge.
MIT Press.
Chandler, A.
1990 Scale and scope: the dynamics 01
industrial capitalismo Cambridge,
Mass.: HarvardIBelknap.
Ciborra, C.
1991 'Allianees as learning experienees:
eooperation, competition and
ehange in high-teeh industries' in
Strategic partnerships and the world
economy. L. Mytelka (ed.), 51-77.
London: Pinter.
Clark, K., R. Hayes, and C. Lorenz
1987 The uneasy alliance: Managing the
productivity-technology dilemma.
Cambridge, Mass.: Harvard Busi-
ness Sehool Press.
392 Mark Dodgson
- - - - - - - - , - - - - - - - - - - - - - _ . _ - - - , - - - - - - - - - - - _ . _ - - - - - _..._•....
Cohen, W., and D. Levinthal
1989 'Innovation and learning: the two
faces of R&D'. The Economic
Journal 99: 569-596.
Corsini, R.
1987 Concise encyclopedia 01 psychology.
New York: Wiley.
Cyert, R., and J. March
1963 A behavioural theory 01 the firmo
Englewood Cliffs. N.J.: Prentice
Hall.
Dodgson, M.
1991a 'Technology learning, technology
strategy and competitive pressures'.
British Journal 01 Management 2/3:
132-149.
Dodgson. M.
1991b The management 01 technological
learning. Berlin: De Gruyter.
Dodgson, M.
1993a 'Learning, trust and technological
collaboration', Human Relations:
\32-149.
Dodgson, M.
1993b Technological collaboration in
industry. London: Routledge.
Dore, R.
1973 British lactory-Japanese lactory.
Berkeley: University of California
Press.
Dore, R.
1986 Taking Japan seriously. London:
Athlone Press.
Dore, R., and M. Sako
1989 How the Japanese learn to work.
London: Routledge.
Dosi, G.
1982 'Technological paradigms and tech-
nological trajectories: A suggested
interpretation of the detenninants
and directions of technical change'.
Research Poliey 2/3: 147-162.
Dosi, G.
1988 'Sources, procedures, and micro-
economic effects of innovation' .
Journal 01 Economic Literature 26:
1120-1171.
Dumbleton. J.
1986 Management 01 high teehnology
R&D. Amsterdam: Elsevier.
Fiol, c., and M. Lyles
1985 'Organizational learning Aeademy
01 Management Review 10/4:
803--813.
Freeman, C.
1982 The economics 01 industrial innova-
tion. London: Pinter.
Freeman, C., and C. Perez
1988 'Structural crises of adjustment:
business cycles and tnvestment
behaviour' in Technieal change and
economic theory. G. Dosi, et aL
(eds.), 38-66. London: Pinter.
Garratt, R.
1987 The learning organization. London:
FontanalCollins.
Hedberg, B.
1981 'How organizations learn and
unlearn' in Handbook 01 organiza-
tional design, Vol. 1. P. Nystrom
and W. Starbuck (eds.), 3-27.
Oxford: Oxford University Press.
Hobday, M.
1990 Telecommunications in developing
countries: the challenge from Brazü.
London: Routledge.
Hull, F.
1988 'Inventions from R&D: organiza-
tion design for efficient research
performance'. Soeiology 22/3:
393-415.
Imai, K-l., 1. Nonaka, and H. Takeuchi
1985 In The uneasy alliance: managing the
productivity-technology dilemma. K.
Clark. R. Hayes and C. Lorenz
(eds.). Cambridge: Harvard Busi-
ness School Press.
Kantcr. R.
1989 When giants learn to lÚlnee. London:
Simon and Schuster.
Kolodny, H.
1980 'Matrix organization designs and
new product success' . Research
Management (September): 29-33.
Levitt, B., and J. March
1988 'Organizational learning'. Annuai
Review 01 Soeiology 14: 319-340.
Loveridge, R., and M. Pitt
1990 The strategic management ol teehno-
logical innovation. Chichester:
Wiley.
Organizational Learning: A Review of Sorne Literatures
Maidique, M., and B. Zirger
1985 'The new product leaming cycle'.
Research Policy 14: 299-313.
Marceau, J.
1992 Reworking the world: organization,
technologies and cultures in com-
parative perspective. Berlín: De
Gruyter.
March, J.
1991 'Exploration and exploitation in
organizational leaming'. Organiza-
tíon Science 2/1: 71-87.
March, J., L. Sproull, and M. Tamuz
1991 'Leaming from samples of one or
fewer'. Organizatíon Science 2/1:
1-13.
Marengo, L.
1992 'Knowledge, coordination and
leaming in an adaptive model of the
firm'. Unpublished D.Phil. disser-
tation, Science Policy Research
Unit, University of Sussex.
Maschlup, F.
1982 Knowledge: its creation, distribution
and economic significance.
Princeton: Princeton University
Press.
Metcalfe, S., and M. Gibbons
1989 'Technology, variety and organiza-
tion' im Research in technological
innovation. Management and policy.
R. Rosenbloom and R. Burgleman
(eds.). New York: JAI Press.
Michael, D.
1973 On learning to plan - and planning
to learn. San Francisco: Jossey-Bass.
Mody, A.
1990 'Leaming through alliances'.
Washington: The World Bank.
Morgan, G.
1986 Images o[ organization. Newbury
Park: Sage.
Mowery, D.
1981 'The emergence and growth of
industrial research in American
manufacturing, 1899-1946'. Mimeo,
Stanford University, Stanford.
Nelson, R. and S. Winter
1982 'An evolutionary theory of econ-
omic change. Cambridge, Mass.:
Belknap Press.
393
Pavitt, K.
1991 'Key characteristics of the large
innovating firm'. British Journal o[
Management 2: 41-50.
Pedler, M., T. Boydell, and J. Burgoyne
1989 'Towards the leaming company'.
Management Education and
Development 20/1: 1-8.
Peters, T., and R. Waterman
1982 In search o[ excellence: lessons [rom
America's best run companies. New
York: Harper and Row.
Prahalad, C. and G. Hame1
1990 'The core competence of the cor-
poration'. Harvard Business Review
(May-June): 79-91.
Pucik, V.
1988a 'Strategic alliances, organizational
leaming, and competitive advan-
tage: the HRM Agenda'. Human
Resource Management 27/1: 77-93.
Pucik, V.
1988b 'Strategic alliances with the
Japanese: implícations for human
resource management' in Coopera-
tive strategies in international busi-
ness. F. Contractor and P. Lorange
(eds.), 487-498. Lexington, Mass.:
Lexington Books.
Rosenberg, N.
1976 Perspectives on technology. Cam-
bridge: Cambridge University Press.
Rosenberg, N.
1982 Inside the black box: technology and
economics. Cambridge: Cambridge
University Press.
Rothwell, R.
1992 'The 5th generation innovation pro-
cess'. R&D Management 22/3:
221-239.
Sako, M.
1992 Prices, quality and trust: how
Japanese and British companies
manage buyer-supplier relations.
Cambridge: Cambridge University
Press.
Schein, E.
1985 Organizational culture and leader-
ship. San Francisco: Jossey-Bass.
394 Mark Dodgson
Senge, P.
1990 'The Ieader's new work: building
learning organizations'. Sloan
Management Review 3211: 7-23.
Simon, H.
1957 Administrative behaviour. New
York: Macmillan.
Simon, H.
1991 'Bounded rationality and organiza-
tional learning'. Organization
Science 2/1: 125-134.
Snow, C., and R. Miles
1983 'The role of strategy in the develop-
men! of a general theory of
organizations' in Advances in
strategy. R. Lamb (ed.). New York:
lAI Press.
Stinchcombe, A.
1990 Information and organizations.
Berkeley: University of California
Press.
Teece, D., G. Pisano, and A. Schuen
1990 'Firm capabilities, resources, and
the concept of strategy'. cce
Working Paper No 90--8, University
of Berkeley, Berkeley.
Twiss, B.
1986 Managing technological innovation.
3rd edn. Landon: Pitman.
von Hippel, E.
1988 The sources of innovatíon. Cam-
bridge: Cambridge University Press.
Whipp, T., R. Rosenfeld, and A.
Pettigrew
1990 Culture and competitiveness: evi-
dence from mature U. K. industries.
Warwick: University of Warwick
Centre for Corporate Strategy and
Change.
Womack, 1., D. Jones, and D. Roos
1990 The machine that changed the world.
New York: Rawson Ass.
Copyright © 2002 EBSCO Publishing